TY  - RPRT
TI  - Scalable Reinforcement Learning via Adaptive Batch Scaling
AU  - Jongchan Park
PY  - 2026
UR  - https://arxiv.org/abs/2605.21557
ID  - 2605.21557
ER  -